In cct. BIOS flashing

[JEWilson]

I have a number of mobos which will not POST.

These are all soldered as opposed to socketed BIOS flash parts.
I suspect the BIOS image is bad or has bad bootblock or a bad IC.

With my POST card;

Mobo 1 - Award
Attempts a BIOS restore, finds the image on diskette, reads
it but does not update the BIOS and simply loops through
the recovery process with no BIOS being restored.
Suspect bad BIOS IC. Reflash it in the first instance.

Mobo 2 - Award
Simply sits a C1 checkpoint. So... it may have an intact bootblock
but cannot load the BIOS image from the IC.

Mobo 3 - Award
The BIOS rdy led is lit and the data LED are flashing
Suspect the BIOS IC is being accessed but the BIOS IC
is damaged, empty or corrupt in some fashion.

Is it safe to flash these in place with a Willem or any other programmer?
I suspect if so, an adapter would be required to do this or a plcc32 socket
rigged to a programmer might do the trick

Any suggestions?

[pfrcom]

Re: In cct. BIOS flashing

If you have any success with these, please post how you did it


I've never had any luck with soldered BIOS chips which have corrupted firmware

If it won't boot into Award boot block recovery, or respond to Ctrl+Home for AMI, it's not worth wasting more time on it in my opinion


In my experience, these recovery tips from the Net DON'T work

1. Shorting address line connections to trigger BIOS data corruption

2. Putting a correctly programmed chip in parallel with the original


I'm not a fan of Gigabyte, the main offenders for soldered BIOS chips, and not impressed with their Dual BIOS either

Don't believe you can trust firmware, to recover from an abnormal event where the original (BIOS) firmware has failed


I believe the new generation of SPI BIOS chips has provision for in circuit reprogramming

[Agent24]

Re: In cct. BIOS flashing

I did manage to reflash a soldered in PLCC32 parallel EEPROM on an Asus CD-RW drive but got hardware error from the programmer when trying the same thing on a motherboard (I used a Willem PCB 5)

I wrote about it here: http://forum.rpc1.org/viewtopic.php?f=13&t=44275

You can see the homebrew adaptor I created

The best way would probably be to get hold of a hot air rework station, and just remove the chips.

Then dump the contents with a programmer, and try to determine if it's corrupt. OR just reflash anyway.

If you went to all that trouble then you should solder a socket to the board before you put the chip back.

[lucky13]

Re: In cct. BIOS flashing

I did manage to reflash a soldered in PLCC32 parallel EEPROM on an Asus CD-RW drive but got hardware error from the programmer when trying the same thing on a motherboard (I used a Willem PCB 5)

I wrote about it here: http://forum.rpc1.org/viewtopic.php?f=13&t=44275

You can see the homebrew adaptor I created

You know, I thought about how to flash my Gigabyte board that would not POST and I have been looking at these programmers and wondered out loud that why there isn't an adaptor to cover over the soldered bios chip and flash away. Your design exactly illustrates my thought. I like it!

Did you write down the p/n of the socket and whatever parts you had to buy?

[Agent24]

Re: In cct. BIOS flashing

I think it has not been invented normally because unlike serial EEPROMs the parallel ones are not designed to support it (in-circuit programming). Probably just chance that it would work on any given board.

I suspect the design of the surrounding circuitry would play some part. It worked for me on the Asus CD-RW but not on an Albatron motherboard. However on the CD-RW drive I had to remove 2 SMD parts to get the socket to fit over the EEPROM. Perhaps this was part of why it worked.

I would test it on some other soldered in PLCC chips I have but they are all FWH/LPC


The design of my adaptor is simple, there are only 4 parts to it and they should be available anywhere.

1) PLCC32 Socket (I used standard type, not SMD) (Jaycar)
2) Prototype board. Similar to Veroboard/Stripboard but without strips. Each hole has a copper pad but none are connected (DSE)
3) SIL Headers. Snap off to 16 pins long, one each side. (DSE)
4) Solid-core ethernet cable. 4 x 8 conductors = 32! Perfect. (DSE)

My adaptor design worked but the implementation is not so good. It was a prototype and I can see there are problems

eg: accidental short circuit is too easy and solid core cable can break easily (however if you DO build one this way, solid core is the only way to do it)

If I built another I would improve on the design by making my own PCBs, and using a ribbon cable and headers to connect the two.

Also, an SMD PLCC socket and/or a handle on the back somehow may help because the one I had did not stay on the EEPROM. It would jump straight off again so I had to hold it down quite hard while I did the programming. Not easy when there's 32 exposed wires in the way and it hurts your fingers.

Remember that the pinout for the PLCC32 will be flipped and changed around. Also, I had to break one corner off the socket to be able to fit it on backwards.

[lucky13]

Re: In cct. BIOS flashing

I remember when I was studying electronics years ago, someone was making these breakout IC clips that you clamp onto the IC and there are individual pins on the top for you to hook up your scope probe. I think something like that may be easier to use except I don't remember the brand nor the proper technical name. That would elimiate the mods that are described by Agent24

[Agent24]

Re: In cct. BIOS flashing

I thought those were only for standard DIL ICs though, not PLCC?

[lucky13]

Re: In cct. BIOS flashing

You might be right...I may have to browse the digikey catalog when I am bored at work....